Problems with expert opinion

Abstract

Cook et al.’s excellent case report and analysis of expert opinion concerning the management of a difficult airway for emergency thyroidectomy raises two important issues [1]. Firstly, the authors conclude that expert opinion varies considerably despite the facts of a case, particularly if the evidence base for best practice is limited, and that this may be problematic when judging third-party performance. However, Stevens, in an equally impressive accompanying editorial [2], correctly points out that since Bolitho, the courts may reject expert opinion if they consider it unreasonable; moreover, the Civil Procedure Rules mandate divulgence of both opinion and published evidence by experts. I would suggest that the implicit problem highlighted in both papers is more to do with the ‘adversarial’ nature of the English and Welsh courts system when hearing civil cases of medical negligence, which is perceived to be as concerned with the gladiatorial contest between opposing lawyers seeking conditional fees as it is with redress through reassurance of future safe practice. The latter would be better served through a more European or coronial-style ‘inquisitorial’ process, aimed at determining the events of a case rather than apportioning blame, an approach that would be more likely to avoid the ‘shame and blame culture of fear’ that Jarman (in an editorial in the same issue [3]) identifies as a major barrier to quality improvement in the National Health Service. A second problem implicit in this article concerns the notion of a ‘best’ technique. Whilst wishing to avoid a relativist ‘all methods are equally valid’ position, I would suggest that the divergence of opinion amongst Cook et al.’s august body of experts represents an underappreciated concept in clinical practice, namely that of operator-dependent risk. Large randomised controlled studies and meta-analyses are viewed as the gold standard of medical practice for their objective determination of the best technique if all other factors (including operator performance) remain equal, but fail to appreciate the context-sensitive, subjective outcomes of individual practitioners (i.e. what is the method that achieves the best outcomes in that individual’s hands under the circumstances that arise?). Expert opinion and quasi-legal guidelines often fail to recognise that insisting one technique is better than others may condemn some anaesthetists to performing techniques with which they are less familiar or that they themselves suspect they are less competent at performing (even if they perform these techniques well and routinely); there may also be an element of risk to patients when an anaesthetist abandons a discredited technique that they perform well whilst they are learning a new technique to the same standard of outcome [4]. Patients want to know ‘what is the risk of this procedure when you’re doing it, doctor’, and it is for this reason that anaesthetists should receive continued encouragement to audit their own practice, in order to be able both to quote operator-dependent risk during consent discussions, and, where their risk compares unfavourably with published risk figures [5], to identify areas for continued professional development.